Ship operation energy consumption prediction method considering actual storm conditions

Abstract

The invention relates to a ship operation energy consumption prediction method considering an actual storm condition. The method comprises the following steps: fusion and data processing of multi-source navigation environment data; the still water resistance and the wave increase resistance are calculated through a program, and the navigational speed-power relation of the ship under the actual wind wave condition is calculated; predicting the power-oil consumption relation of the ship according to the route segments: based on the navigational speed-power relation of the ship under the actual storm condition, establishing a prediction agent model of the power-oil consumption relation of the ship according to the historical oil consumption data of the ship; and ship operation energy consumption prediction: combining the annual route segment information of the ship, predicting the agent model based on the branch route segment ship power-oil consumption relation, thereby obtaining the oil consumption of each route segment of the ship in the whole year, and calculating the annual operation energy consumption of the ship. According to the method, the prediction precision is remarkably improved, the generalization ability and the real-time response speed of the model are enhanced, and powerful support is provided for energy conservation and emission reduction of ships.

Description

Technical Field

[0001] The present invention relates to the technical field of ship energy efficiency, and in particular to a method for predicting ship operating energy consumption considering actual wind and wave conditions. Background Art

[0002] With the intensification of the global energy crisis and the increasing severity of environmental pollution problems, as a major means of transportation, the energy consumption and emissions of ships have received wide attention.

[0003] During the actual operation of a ship, it is affected by natural conditions such as wind and waves, which will significantly change the ship's energy consumption. In recent years, the International Maritime Organization (IMO) has been increasingly strict with the requirements for ship energy efficiency. Therefore, developing a method for predicting ship operating energy consumption that can consider actual wind and wave conditions is of great significance for improving ship energy efficiency, reducing operating costs, and reducing environmental pollution. In recent years, with the rising energy prices and the increasingly prominent environmental problems, how to effectively reduce the operating costs of ships and reduce their impact on the environment has become the focus of attention in the industry.

[0004] Previous ship energy efficiency predictions often relied on theoretical calculations and tests under ideal conditions, but ignored the impact of the complex and variable wind and wave environment on ship energy consumption in the real marine environment. The impact of wind and waves on ship energy consumption will affect the prediction of ship energy consumption, and thus affect the accuracy of energy consumption prediction.

[0005] The existing technologies mainly include an energy consumption estimation method based on empirical formulas and a simulation analysis method based on physical models. The former usually relies on a large amount of actual ship test data and estimates the ship's energy consumption under different working conditions through an empirical formula obtained by fitting; the latter is to predict the energy consumption performance of the ship under specific conditions by establishing a detailed hydrodynamics model and combining meteorological forecast information. However, these methods are either too simplified to fully reflect the impact of the complex marine environment on ship performance, or have a huge amount of calculation and are difficult to achieve rapid response in practical applications.

[0006] The existing ship energy consumption prediction methods have improved the accuracy of ship energy consumption prediction to a certain extent, but there are still the following deficiencies: Limited accuracy: Most of the existing models are based on test data under static conditions and do not fully consider the impact of the dynamically changing wind and wave environment on ship energy consumption, resulting in a certain deviation between the prediction results and the actual situation. Poor real-time performance: Although the method based on physical models can theoretically provide a high prediction accuracy, due to the need for a large amount of complex calculations, it is difficult to meet the requirements of real-time decision-making. High cost: High-precision energy consumption prediction methods often rely on expensive data acquisition equipment and high-performance computing resources, which are high-cost expenditures for small and medium-sized shipping enterprises.

[0007] For this reason, we propose a method for predicting the operating energy consumption of ships considering actual wind and wave conditions. Summary of the Invention

[0008] The applicant provides a method for predicting the operating energy consumption of ships considering actual wind and wave conditions, aiming at the shortcomings in the above-mentioned existing production technologies. This method not only significantly improves the prediction accuracy, but also enhances the generalization ability and real-time response speed of the model, providing strong support for the energy conservation and emission reduction of ships.

[0009] The technical solution adopted by the present invention is as follows:

[0010] A method for predicting the operating energy consumption of ships considering actual wind and wave conditions, comprising the following steps:

[0011] Fusion and data processing of multi-source navigation environment data: The navigation environment data includes the longitude and latitude coordinate points of the ship's route, main engine power, speed, heading, wind speed, wind direction, current speed, current direction, and sea conditions;

[0012] Calculation of still water resistance and wave added resistance: Calculate the still water resistance and wave added resistance through a program, and then calculate the speed-power relationship of the ship under actual wind and wave conditions;

[0013] Prediction of the power-fuel consumption relationship for each voyage segment of the ship: Based on the speed-power relationship of the ship under actual wind and wave conditions and combined with the ship's historical fuel consumption data, establish a prediction proxy model for the power-fuel consumption relationship for each voyage segment of the ship;

[0014] Prediction of the operating energy consumption of the ship: Combine the annual voyage information of the ship and, based on the prediction proxy model for the power-fuel consumption relationship for each voyage segment of the ship, obtain the fuel consumption of each voyage segment of the ship throughout the year, and then calculate the annual operating energy consumption of the ship; The operating energy consumption of the ship refers to the ship's operating carbon intensity index CII, and its formula is:

[0015]

[0016] where FC j is the total mass of type j fuel consumed within the calendar year, in grams;

[0017] CF j is the fuel mass to CO2 conversion coefficient for type j fuel;

[0018] C represents the carrying capacity of the ship;

[0019] D t represents the total navigation distance reported in IMO DCS, in nautical miles.

[0020] It is further characterized in that:

[0021] For bulk carriers, tankers, container ships, gas carriers, LNG ships, general cargo ships, refrigerated cargo ships and combination carriers, deadweight tons shall be used as the carrying capacity.

[0022] For luxury cruise ships, roll-on / roll-off cargo ships, roll-on / roll-off passenger ships, gross tons shall be used as the carrying capacity.

[0023] A method for multi-source navigation environment data fusion and processing includes the following steps:

[0024] Obtaining and collecting multi-source navigation environment data: The sources of multi-source navigation environment data include ship navigation measured data, ship noon reports, and global ocean meteorological data. Collect and screen the longitude and latitude coordinate points of the ship's route, main engine power, speed, course, loading capacity, wind speed, wind direction, current speed, current direction, and sea condition data among them;

[0025] Data cleaning: Based on the collected multi-source navigation environment data set, perform data cleaning, including the unification of the data sampling period and the identification of abnormal data;

[0026] The data sampling frequencies of different sources and dimensions in the multi-source navigation data set are inconsistent. It is necessary to uniformly process the sampling period before modeling. The formula for unifying the sampling period is as follows:

[0027]

[0028] Where a is the sampling period, in seconds;

[0029] Y i is the i-th data in the data set after the unified processing of the sampling period;

[0030] is the sum of the data from the i-th second to the i + a-th second in the original data set;

[0031] Consistent transformation of spatial data: The wind speed and current speed in the multi-source navigation environment data set are often the wind speed and current speed relative to the ground. To ensure the accuracy of subsequent model training, it is necessary to perform a consistent transformation on the spatial data after the above data cleaning, including wind speed and current speed.

[0032] On the basis of the consistent processing of the sampling period, the common outlier processing methods in data processing are used to correct abnormal data, including but not limited to missing value processing, data deduplication, and data interpolation.

[0033] For the transformation of wind speed, where Vwind represents the wind speed relative to the ground in the multi-source navigation environment data set, α is the angle between the ship's wind direction and course, Vwindh is the component of the wind speed in the ship's bow direction, and Vwindc is the component of the wind speed in the ship's lateral direction. The calculation formula is as follows:

[0034] Vwindh = V wind × cosα

[0035] V windc = V wind × sinα

[0036] The conversion calculation formula for flow velocity is as follows:

[0037] V c = V g - V s

[0038] Wherein, V c is the flow velocity; V g is the ship's speed over the ground; V s is the ship's speed through the water.

[0039] The method for predicting the relationship between ship power and fuel consumption in a voyage section includes the following steps:

[0040] Normalization processing: To improve the accuracy of modeling, normalization processing needs to be carried out before modeling, and normalization and feature selection are performed on the processed data set;

[0041] Establishment of the ship power - fuel consumption relationship prediction model for each voyage section: Considering the ship operation and operation characteristics, it is established in units of voyage sections, and an artificial neural network based on time series is used. The ship navigation wind and wave conditions, route information, and the speed - power relationship of the ship under actual wind and wave conditions are used as inputs, and the total fuel consumption within the ship voyage section is used as the output to construct a prediction model.

[0042] The normalization methods include min - max normalization and Z - Score normalization.

[0043] The wind and wave conditions of ship navigation include: sea conditions, as well as the wind speed and flow velocity after multi - source navigation environment data fusion, processing, and normalization.

[0044] The route information of ship navigation includes: the longitude and latitude coordinate points after multi - source navigation environment data fusion, processing, and normalization.

[0045] The beneficial effects of the present invention are as follows:

[0046] The present invention can accurately reflect the influence of actual wind and wave conditions on ship performance. By constructing a multi - variable non - linear prediction model, combining real - time obtained marine meteorological environment parameters such as wind speed, wave height, and flow direction, as well as operation state information such as the ship's speed and loading condition, it realizes the accurate prediction of ship energy consumption under complex sea conditions. Compared with the prior art, the present invention not only significantly improves the prediction accuracy, but also enhances the generalization ability and real - time response speed of the model, providing strong support for ship energy conservation and emission reduction. Description of the Drawings

[0047] Figure 1 Schematic flow diagram of the energy consumption prediction method of the present invention.

[0048] Figure 2 Schematic flow diagram of the method for multi-source navigation environment data fusion and processing of the present invention.

[0049] Figure 3 Schematic diagram of the spatial consistency transformation of wind speed data of the present invention.

[0050] Figure 4 Schematic diagram of the neural network of the present invention.

[0051] Figure 5 Schematic diagram of the training criterion of the present invention.

[0052] Figure 6 Schematic diagram of the prediction of the power-fuel consumption relationship of ships in different voyage segments of the present invention. Detailed implementation manners

[0053] The following will describe the detailed implementation manners of the present invention with reference to the accompanying drawings.

[0054] As Figures 1 - 6 shown, a ship operation energy consumption prediction method considering actual wind and wave conditions includes the following steps:

[0055] Fusion and data processing of multi-source navigation environment data: The navigation environment data includes ship route longitude and latitude coordinate points, main engine power, ship speed, course, wind speed, wind direction, current speed, current direction, and sea conditions; its sources include ship navigation measured data, ship noon report data, and global ocean meteorological data; the data processing method is the fusion of multi-source data, data cleaning, and spatio-temporal data unification;

[0056] Calculation of still water resistance and wave added resistance: Calculate the still water resistance and wave added resistance through a program, and then calculate the speed-power relationship of the ship under actual wind and wave conditions;

[0057] Prediction of the power-fuel consumption relationship of ships in different voyage segments: Based on the speed-power relationship of the ship under actual wind and wave conditions, combined with the ship's historical fuel consumption data, establish a prediction proxy model for the power-fuel consumption relationship of ships in different voyage segments;

[0058] Ship operation energy consumption prediction: Combining the ship's annual voyage information, based on the prediction proxy model of the power-fuel consumption relationship of ships in different voyage segments, obtain the fuel consumption of each voyage segment of the ship throughout the year, and then calculate the annual operation energy consumption of the ship; The ship operation energy consumption refers to the ship operation carbon intensity index CII, and its formula is:

[0059]

[0060] where, FCj is the total mass (grams) of type J fuel consumed within a calendar year;

[0061] CF j is the fuel mass to CO2 conversion coefficient of type J fuel;

[0062] C represents the carrying capacity of the ship: for bulk carriers, liquid cargo ships, container ships, gas carriers, LNG ships, general cargo ships, refrigerated cargo ships and combination carriers, deadweight tons (DWT) should be used as the carrying capacity, and for luxury cruise ships, roll-on / roll-off cargo ships (vehicle carriers), roll-on / roll-off ships and roll-on / roll-off passenger ships, gross tons (GT) should be used as the carrying capacity;

[0063] D t represents the total navigation distance (nautical miles) reported in IMO DCS.

[0064] As Figure 2 shown, the method for multi-source navigation environment data fusion and processing includes the following steps:

[0065] Obtaining and collecting multi-source navigation environment data: The sources of multi-source navigation environment data include ship navigation measured data, ship noon reports, and global ocean meteorological data. Collect and screen the ship's route longitude and latitude coordinate points, main engine power, speed, course, loading volume, wind speed, wind direction, current speed, current direction, and sea condition data among them.

[0066] Data cleaning: Based on the collected multi-source navigation environment data set, perform data cleaning, including the unification of data sampling periods and the identification of abnormal data;

[0067] The data sampling frequencies of different sources and dimensions in the multi-source navigation data set are inconsistent, and the sampling period needs to be uniformly processed before modeling. The formula for sampling period unification is as follows:

[0068]

[0069] where a is the sampling period, in seconds;

[0070] Y i is the i-th data in the data set after the unified processing of the sampling period;

[0071] is the sum of the data from the i-th second to the i + a-th second in the original data set.

[0072] On the basis of the unified processing of the sampling period, the common outlier processing methods in data processing are used to correct the abnormal data, including but not limited to missing value processing, data deduplication, and data interpolation, etc.

[0073] Spatial data consistency transformation: The wind speed and flow velocity in the multi-source navigation environment dataset are often the wind speed and flow velocity relative to the ground. To ensure the accuracy of subsequent model training, it is necessary to perform consistency transformation on the spatial data after the above data cleaning, including wind speed and flow velocity.

[0074] As Figure 3 shown, for the transformation of wind speed, where Vwind represents the wind speed relative to the ground in the multi-source navigation environment dataset, α is the angle between the ship's wind direction and the course, Vwindh is the component of the wind speed in the ship's head direction, and Vwindc is the component of the wind speed in the ship's lateral direction. The calculation formulas are as follows:

[0075] V windh = V wind × cosα

[0076] V windc = V wind × sinα

[0077] The calculation formula for the transformation of flow velocity is as follows:

[0078] V c = V g - V s

[0079] Among them, V c is the flow velocity; V d is the ship's speed relative to the ground; V s is the ship's speed through the water.

[0080] Method for predicting the power-fuel consumption relationship of ships in a voyage section, including the following steps:

[0081] Normalization processing: To improve the accuracy of modeling, it is necessary to perform normalization processing before modeling. Normalization and feature selection are performed on the processed dataset. The normalization methods include but are not limited to min-max normalization, Z-Score normalization, etc.

[0082] As Figure 4 , Figure 5 shown, establishing a prediction model for the power-fuel consumption relationship of ships in each voyage section: Considering the ship's operation and operation characteristics, it is established in units of voyage sections. An artificial neural network based on time series is used. The ship's navigation wind and wave conditions, route information, and the speed-power relationship of the ship under actual wind and wave conditions are used as inputs, and the total fuel consumption within the ship's voyage section is used as the output to construct a prediction model.

[0083] The wind and wave conditions of ship navigation include: sea conditions, as well as the wind speed and flow velocity after multi-source navigation environment data fusion, processing, and normalization processing.

[0084] The route information of ship navigation includes: the longitude and latitude coordinate points after multi-source navigation environment data fusion, processing, and normalization.

[0085] As shown in Table 1, the input and output parameters and their relationships of the prediction model for the ship power-fuel consumption relationship by voyage segment are presented.

[0086] Table 1 Input and output parameters and their relationships of the prediction model for the ship power-fuel consumption relationship by voyage segment

[0087] Name Conform to Physical quantity Category Time step range Wind and wave conditions E Sea state, wind speed, flow velocity Input quantity [t, t + n] Route information L Latitude and longitude coordinate points Input quantity [t, t + n] Ship output power P Ship output power Input quantity [t, t + n] Ship fuel consumption F Ship fuel consumption Output quantity t + n

[0088] Among them, the ship output power is obtained from the speed-power relationship of the ship under actual wind and wave conditions through an independently developed program by the China Ship Scientific Research Center.

[0089] Such as Figure 6 shown, the prediction model for the ship power-fuel consumption relationship by voyage segment is established. Considering within a single voyage segment, the input and output data are divided into several time units. Each time unit starts from time t, and the value of the output at time t + n is determined by all the input quantities from time t to time t + n.

[0090] It can accurately reflect the influence of actual wind and wave conditions on ship performance. By constructing a multi-variable non-linear prediction model and combining real-time obtained marine meteorological environment parameters such as wind speed, wave height, and flow direction, as well as operation state information such as ship speed and loading condition, the accurate prediction of ship energy consumption under complex sea conditions is realized. Compared with the existing technology, the present invention not only significantly improves the prediction accuracy, but also enhances the generalization ability and real-time response speed of the model, providing strong support for ship energy conservation and emission reduction.

[0091] The above description is an explanation of the present invention, not a limitation of the invention. The scope defined by the present invention can be seen in the claims, and any form of modification can be made within the protection scope of the present invention.

Claims

1. A method for predicting ship operation energy consumption considering actual wind and wave conditions, characterized in that: The steps include: Fusion and data processing of multi-source navigation environment data: Navigation environment data include the latitude and longitude coordinates of the ship's route, main engine power, speed, heading, wind speed, wind direction, current speed, current direction, and sea conditions; Calculation of still water resistance and wave resistance: Calculate still water resistance and wave resistance through the program, and then calculate the speed-power relationship of the ship under actual wind and wave conditions; Prediction of ship power-fuel consumption relationship by voyage section: Based on the speed-power relationship of the ship under actual wind and wave conditions and combined with the historical fuel consumption data of the ship, an agent model for predicting the power-fuel consumption relationship of the ship by voyage section is established; Ship operation energy consumption prediction: Combined with the ship's annual voyage information, based on the ship power-fuel consumption relationship prediction agent model for each voyage segment, the fuel consumption of the ship in each voyage segment throughout the year is obtained, and the ship's annual operating energy consumption is calculated; the ship's operating energy consumption refers to the ship's operating carbon intensity index CII, and its formula is: Among them, FC j is the total mass of type J fuel consumed in the calendar year, in grams; CF j is the fuel mass and CO2 conversion factor of type j fuel; C represents the carrying capacity of the ship; D t Indicates the total distance sailed as reported in the IMO DCS in nautical miles.

2. A method for predicting ship operation energy consumption considering actual wind and wave conditions as claimed in claim 1, characterized in that: For bulk carriers, tankers, container ships, gas carriers, LNG carriers, general cargo ships, refrigerated cargo ships and combination ships, deadweight tonnage should be used as the carrying capacity.

3. A method for predicting ship operation energy consumption considering actual wind and wave conditions as claimed in claim 1, characterized in that: For luxury cruise ships, ro-ro cargo ships, ro-ro cargo ships and ro-ro passenger ships, gross tonnage should be used as the carrying capacity.

4. A method for predicting ship operation energy consumption considering actual wind and wave conditions as claimed in claim 1, characterized in that: The method for fusing and processing multi-source navigation environment data comprises the following steps: Acquisition and collection of multi-source navigation environment data: The sources of multi-source navigation environment data include ship navigation measured data, ship midday report data, and global marine meteorological data. The ship route longitude and latitude coordinates, main engine power, speed, heading, load, wind speed, wind direction, current speed, current direction, and sea condition data are collected and screened; Data cleaning: Based on the collected multi-source navigation environment data sets, data cleaning is performed, including consistency of data sampling cycles and identification of abnormal data; The sampling frequencies of data from different sources and dimensions in the multi-source navigation data set are inconsistent. The sampling period needs to be unified before modeling. The formula for unifying the sampling period is as follows: Where a is the sampling period in seconds; Y i The i-th data in the data set after unified processing of the sampling period; is the sum of the data from the i-th second to the i+a-th second in the original data set; Consistency conversion of spatial data: The wind speed and flow speed in the multi-source navigation environment data set are often the wind speed and flow speed over the ground. In order to ensure the accuracy of subsequent model training, it is necessary to perform consistency conversion on the spatial data after the above data cleaning, including wind speed and flow speed.

5. A method for predicting ship operation energy consumption taking into account actual wind and wave conditions as claimed in claim 4, characterized in that: On the basis of the consistent processing of the sampling period, the abnormal data are corrected by using the outlier processing methods commonly used in data processing, including but not limited to missing value processing, data deduplication and data interpolation.

6. A method for predicting ship operation energy consumption taking into account actual wind and wave conditions as claimed in claim 5, characterized in that: For the conversion of wind speed, Vwind represents the ground wind speed in the multi-source navigation environment dataset, α is the angle between the ship's wind direction and the heading, Vwindh is the component of the wind speed at the bow of the ship, and Vwindc is the component of the wind speed at the side of the ship. The calculation formula is as follows: V windh =V wind ×cosα V windc =V wind ×sinα The conversion calculation formula for flow rate is as follows: V c =V g -V s Among them, V c is the flow rate; V g V is the ship's speed over the ground; s The ship's speed through water.

7. A method for predicting ship operation energy consumption considering actual wind and wave conditions as claimed in claim 1, characterized in that: The method for predicting the relationship between power and fuel consumption of a voyage ship comprises the following steps: Normalization: In order to improve the accuracy of modeling, normalization is required before modeling, and normalization and feature selection are performed on the processed data set; Establishment of a prediction model for the relationship between ship power and fuel consumption in different voyage sections: Taking into account the ship's maneuvering and operating characteristics, it is established in sections, using a time-series-based artificial neural network. The ship's sailing wind and wave conditions, route information, and the ship's speed-power relationship under actual wind and wave conditions are taken as input, and the total fuel consumption of the ship in the voyage section is taken as output to construct a prediction model.

8. A method for predicting ship operation energy consumption taking into account actual wind and wave conditions as claimed in claim 7, characterized in that: Normalization methods include maximum value normalization and Z-Score normalization.

9. A method for predicting ship operation energy consumption taking into account actual wind and wave conditions as claimed in claim 7, characterized in that: The wind and wave conditions for ship navigation include: sea conditions, wind speed and current speed after fusion, processing and normalization of multi-source navigation environment data.

10. A method for predicting ship operation energy consumption taking into account actual wind and wave conditions as claimed in claim 7, characterized in that: The ship's navigation route information includes: longitude and latitude coordinate points after fusion, processing and normalization of multi-source navigation environment data.

Images